一种新的自相似流量模型的网络性能分析

Analysis of Network Performance for a New Self-similar Traffic Model

网络流量的自相似性给流量控制和网络资源的管理带来了新的挑战。基于传统流量模型的网络性能结论和网络设计策略不能完全适用于具有自相似流量的高速网络中,采用较符合实际流量特征的、并具有自相似特性的随机过程——随机相位正弦波作为队列服务系统的输入流量模型,运用排队理论推导和分析了这种新的流量模型的网络性能,然后探讨了改善自相似流网络性能的流量控制方法和设想。

The characteristic of self-similarity has a challenge to traffic control and network resource management. Both the conclusion of network performance and the strategy of network design based on Poisson modeling are not quite suitable for the networks of self-similarity. A sine wave with random phase is used, a stochastic process that takes on the characteristic of self-similarity, as the input tragic model of the queuing system. Based on queuing theory, the queuing performance of this new traffic model deduces and analyzes, then discusses the methods and suggestions used for traffic control in improving network performance for self-similar traffic.